UNIVERSIDAD NACIONAL DE INGENIERIA. FACULTAD DE INGENIERIA QUIMICA. AREA ACADEMICA DE INGENIERIA QUIMICA
DISEO DE PLANTAS CASO: EVALUACIN D ETECNOLOGIAS PARA LA DESHIDRATACIN DE ETANOL PARA LA MANUFACTURA
D EALCOHOL ABSOLUTO GENERALIDADES: Ud. Forma parte de un grupo de ingenieros de procesos que ha recibido
el encargo de realizar una evaluacin econmica preliminar de tecnologas para la obtencin de alcohol absoluto para uso como
combustible a partir de mezclas de etanol con agua. Se trata de determinar el costo mximo que podra pagarse por el alcohol
hidratado de manera de poder emplearlo como combustible en mezclas con gasolina. Memo PI525-001
0709.2001 A: Ingeniero de Procesos De: Jefe Ingeniera de Procesos: Como es conocido el etanol es
el alcohol utilizado para limpiar heridas y prevenir infecciones. Este alcohol fue utilizado por Henry Ford, cuando dise el
primer modelo de automvil T, como combustible. El etanol se obtiene de varias fuentes, entre ellas destacan la caa de
azcar, el maz, papas, madera o cualquier otra sustancia que contenga azcar o ltimamente celulosa que pueda ser fermentada.
Lo que se hace es dejar que el azcar de estos productos sea fermentado por levaduras (unos microbios que comen azcar y desechan
alcohol). Otra manera de obtener etanol es a partir del etano. El etano es un gas que surge de la tierra cuando se extrae
petrleo o gas natural; este etano puede ser procesado por una planta petroqumica para obtener etanol. Como es muy barato
producir el etanol con las levaduras se prefiere utilizar el etano que surge con el petrleo para la manufactura de plsticos.
El etanol es una fuente renovable de energa. Para utilizar el etanol como combustible en motores de combustin interna
se tienen varias maneras; ya sea llenar todo el tanque (el 100%) con etanol, o bien, colocar 85% de etanol y lo que sobra
de gasolina; a esta mezcla se le llama E85. El motor de los automviles que utilizan gasolina pueden tambin funcionar con E85,
as que no hay que comprar otro coche para poder utilizar combustibles alternativos. El etanol al 100% da menos energa
que la mezcla E85, ya que el poco de gasolina que se encuentra en E85 tambin se quema, y la gasolina tiene ms energa que el
etanol. Otra forma ms antigua es el empleo del denominado gasohol que es una mezcla de 80% de gasolina y 20% de etanol
anhidro o alcohol absoluto. Esta forma de empleo est muy extendida en Norteamrica, donde el etanol anhidro incluso est reemplazando
al MTBE como un compuesto mejorador de octanos. Hoy en da muchos pases ya tienen en sus ciudades automviles utilizando
etanol como combustible. El pas que tiene ms trayectoria en esto es Brasil. Durante mucho tiempo el obtener alcohol de
la biomasa no fue competitivo con el empleo de combustibles derivados del petrleo. Sin embargo los desarrollos de la bioqumica
han sido tan impresionantes desarrollando una ruta enzimtica para la produccin de alcohol hidratado a partir de la celulosa
que en el ao 2,000 se inaugur la primera planta de demostracin con dinero del Departamento de Energa de los Estados Unidos.
Conjuntamente con los desarrollos biotecnolgicos ha habido un enorme desarrollo en las tecnologas para producir etanol
anhidro (alcohol absoluto) a partir del alcohol hidratado. En el ao 2004 quedar prohibido el empleo de plomo tetra-etlico
en las gasolinas peruanas con lo que se presenta la oportunidad de implantar el empleo de mezclas alcohol absoluto gasolina
primaria (gasohol) Su asignacin ser plantear dos esquemas de procesos para la obtencin del alcohol absoluto de modo
que pueda ser empleado en las mezclas alcohol absoluto gasolina a partir del ao 2004. Ud. Deber encontrar cual es el valor
mximo que se podra pagar por el alcohol en la puerta de la destilera para que pudiera emplearse en la manufactura de gasolina
de A octanos a partir del pool de gasolinas mostrado. Debido a preocupaciones ambientales Uds. No podrn emplear el
la tecnologa de la destilacin con benceno para deshidratar el etanol. Como el empleo del alcohol tendr una base ecolgica
Ud. puede considerar que la fraccin de etanol anhidro contenido en la gasolina final no pagar ni impuesto al rodaje ni el
impuesto selectivo al consumo. Como los precios de los combustible varan en el tiempo, Uds. Debern ser muy cuidadosos
en efectuar un anlisis de sensibilidad al precio de los componentes de la gasolina y a los valores relativos de ellos.
Memo PI525-002 07.09.01 A: Ingeniero de Procesos De: Jefe Grupo Evaluacin
tecnologas Nuestro grupo ha revisado las diversas tecnologas para la deshidratacin del etanol. En principio el agua
no puede ser eliminada totalmente del etanol por la presencia de un azetropo mnimo. El etanol puede ser deshidratado mediante
destilacin azeotrpica y/o destilacin extractiva. La destilacin extractiva es el mtodo por el cual se separan compuestos
con puntos de ebullicin cercanos destilando en una columna de platos con la presencia de un lquido o una mezcla de lquidos,
estos lquidos tendrn un punto de ebullicin superior al de los componentes que est siendo separados. El agente extractivo se
aade cerca del tope de la columna y cae dentro de la columna alterando la volatilidad relativa de los componentes que van
a ser separados. Cuando los compuestos a ser separados forma azetropos los agentes extractivos harn que ellos ebullan
separadamente y permitirn una adecuada separacin, que no sera posible sin la presencia dela gente extractivo. En
el fondo de la columna el componente menos voltil y el agente extractivo son retirados. Esto son finalmente separados entre
ellos por otra columna de destilacin o por enfriamiento y separacin de fases o por extraccin por solventes. La separacin
del alcohol del etanol es uno de los problemas tcnicos mas antiguos de la industria de procesos. Ud deber presentar
un reporte en 08 das de las tecnologas disponibles para deshidratar etanol Memo PI525-003
14.09.01 A: Ingeniero de Procesos De: Jefe Grupo Evaluacin tecnologas Ud. Deber preparar su
diseo para las condiciones adjuntas GRUPO CAPACIDAD DESTILERA (GAL/DIA) CONCENTRACIN ALCOHOL CRUDO(% ) TIPO GASOLINA
BASE OCTANAJE GASOLINA BASE OCTANAJE GASOHOL 01 10,000 10.0 PRIMARIA 71.0 84 02 11,000 10.5 PRIMARIA 71.5 84 03
12,000 11.0 PRIMARIA 72.0 84 04 13,000 11.5 PRIMARIA 72.5 84 05 14,000 12.0 PRIMARIA 73.0 84 06 15,000 12.5 PRIMARIA
73.5 84 07 16,000 13.0 PRIMARIA 74.5 84 08 17,000 13.5 PRIMARIA 75.0 84 09 18,000 14.0 PRIMARIA 75.5 84 10
19,000 14.5 CRAQUEDA 90.0 95 11 20,000 15.0 CRAQUEDA 91.0 95 12 19,500 15.5 CRAQUEDA 91.5 95 13 18,500 16.0 CRAQUEDA
92.0 97 14 17,500 16.5 CRAQUEDA 90.5 95 15 16,500 17.0 CRAQUEDA 92.5 97 16 15,500 17.5 CRAQUEDA 93.0 97 17
14,500 18.0 CRAQUEDA 93.5 97 18 13,500 18.5 CRAQUEDA 90.5 95 Se ha decidido que el esquema bsico de Procesos
corresponder a una Destilacin Azeotrpica con Pentano como agente azeotrpico. El proceso alterno contra el cual Ud. Deber
validar su diseo deber ser alguna de las tecnologas de la destilacin extractiva Unusual Thermodynamic Properties
and Visualization - New Keys For Process Design Leo C. Kobylka1 and Marco A. Satyro2* 1 Hyprotech Ltd. 2
Hyprotech Ltd. and Department of Chemical and Petroleum Engineering, University of Calgary Author for correspondence
Abstract The continuous development of thermodynamic models provides us with a basic framework for modeling chemical
processes. Process engineers usually relate to the thermodynamic models by using a process simulator and analyzing just a
few points of the composition, pressure and temperature space in great detail. Unfortunately, this very fine grained view
of the thermodynamic space does not allow us to appreciate the way a thermodynamic property behaves in the whole space. As
a result, a wealth of information potentially available from the basic thermodynamic model is not used. In this paper,
we present how thermodynamics can be used to provide us with deep insight on the behavior of a mixture by studying azeotropism,
residue curve maps and thermal maps. These methods give us a visual representation of the thermodynamic model over the whole
component space, aiding in both model understanding and model development. Introduction In the last fifteen years,
we have witnessed a great effort from the academic and industrial community aimed towards the creation of more rigorous thermodynamic
models, or towards the adaptation of simpler models (like cubic equations of state) for more challenging problems. In the
same period, the chemical industry has driven the process simulation companies to provide very detailed thermophysical properties
for process streams. Although the reasons for developing very detailed thermophysical information is easy to understand from
an equipment design point of view, some fundamental aspects linked to the thermodynamic modeling of a mixture tend to be pushed
to a second plane. In our strive to create new and better thermodynamic models, we usually look at just a handful of thermodynamic
properties for pure components and mixtures. Satyro and Trebble (1994) have shown that a thermodynamic model with good characteristics
for VLE prediction can lead to erroneous results for the calculation of derived thermodynamic properties. More importantly,
thermodynamic models are usually tested against binary data only, and we tend to forget that in actual processes, very rarely
one will encounter a binary mixture. The problem for the process or design engineer, faced with the task of understanding
the whole process and expected to pinpoint problems during operation or prevent problems before the plant is actually built,
is daunting. The ability of focusing in (to gather enough information about the unit operation of interest) and out (to appreciate
the overall impact of local decisions) is usually hard to exercise, coming to being only after many years of experience. Therefore,
if we can find ways of presenting the complete thermodynamic surface to the engineer, we should be able to offer a panoramic
view that is abstracted from the actual unit operations used in the process. This overall should be able to show us both inconsistent
models as well as thermodynamic barriers for the operation of chemical processes. Therefore, providing this thermodynamic
"road map" to the engineer can enhance not only productivity, but suggest new ways to solve problems. One important
aspect of modeling is the quality of the physical properties and interaction parameters being used. This fundamental problem
has been analyzed by several authors (Mathias and Copeman (1994), Satyro and Marsh (1994)) and we will not reiterate on its
importance. From the point of view used in this paper, a reasonably adequate thermodynamic model is available to model a mixture.
Our objective is to learn as much as possible from this accepted thermodynamic model. Azeotropism and Residue Curve Maps
Azeotropes are important not only as thermodynamic barriers for separation using distillation (for example, isopropanol
and water mixtures) but also as thermodynamic possibilities to separate mixtures which present a lower boiling azeotrope (like
separating ethanol and water using cyclohexane). However, the presence of azeotropes, especially saddle azeotropes in multicomponent
mixtures, can produce results that are difficult to interpret. From a research point of view, the estimation of azeotropes
based on a given thermodynamic model can suggest an experimental program to gather additional data and further validate the
model predictions. More attention can be paid to specific regions of the composition space, resulting in an optimization of
the time necessary to gather the data really crucial to model the process. Recently, Fidkowski, Malone and Doherty (1993)
proposed a method to compute all the azeotropes in a multicomponent mixture by using a homotopy method together with an arc
length continuation. This method, combined with the residue curve map concept discussed by Doherty and Perkins (1978a,b,1979a)
allows us to see structural relationships in the composition space such as distillation boundaries and distillation regions.
Simply stated, a residue curve is the phase plane for the liquid composition in an isobaric (or isothermal) open evaporation.
The equations describing the process are analogous to the equations describing a batch distillation (or an isothermal evaporation
process driven by a variable pressure) and were discussed by Doherty and Perkins (1978a). A residue curve map has several
interesting characteristics: 1. The residue curves always point in the direction of increasing temperature. 2.
Pure components and azeotropes define fixed points in the map. 3. Azeotropes define boundaries in the composition space,
signaling different qualitative behavior depending on where we are in the space. 4. Boundaries define distillation regions,
which can not be crossed by simple stage-by-stage distillation. 5. Residue curves are equivalent to the composition profile
one would get running a distillation tower at fixed pressure and infinite reflux. The biggest benefit that a residue curve
map provides is that it is a qualitative description of the composition profile that we would observe in an actual distillation
tower. For example, the residue curve map for an almost ideal system of n-Pentane, n-Hexane, and n-Heptane at 1 bar is shown
in Figure 1. Note that just by looking at the residue curve map, we can see that there are no distillation boundaries: pure
n-Pentane is an unstable node, pure n-Hexane is a saddle node and pure n-Heptane is a stable node. The residue curves show
the path traced in the composition space by the liquid in the batch distillation still. Note that the arrows point towards
the direction of increasing temperature (corresponding to the n-Heptane residue). Figure 1: Residue Curve Map for
n-C5, n-C6 and n-C7 at 1 bar. Now lets analyze a ternary mixture with binary azeotropes. The residue curve maps for Acetone,
Methanol and Water at 1 and 6 bar are shown in Figures 2 and 3 respectively. The property package used for both of these cases
is UNIQUAC for the liquid phase and Ideal Gas for the vapor phase. Figure 2: Residue Curve Map for Acetone,
Methanol, and Water at 1 bar . Figure 3: Residue Curve Map for Acetone, Methanol, and Water at 6 bar. First
of all, note the azeotrope between Acetone and Methanol in the maps and how the qualitative behavior of the residue curve
maps change depending on the position relative to the binary azeotrope. For feed mixtures sufficiently poor in water, we can
approach the methanol saddle point, whereas for feed mixtures rich in water, we cannot approach the methanol saddle point.
Now notice the extreme change in the fabric of the composition space when changing the pressure from 1 to 6 bars. A new
azeotrope between Acetone and Water is formed and a distillation boundary is defined. Also note how the azeotrope between
Acetone and Methanol has moved and that there are no residue curves going from the unshaded region towards the shaded region.
The appearance of the distillation boundary gives rise to two different structures in the composition space, and feed mixtures
in one region can not give products in a different region by simple distillation. It can be shown that the bifurcation pressure
for the Acetone and Water azeotrope is around 3 bar. More importantly, from an operational point of view, a tower with
a feed consisting of 0.4 mole fraction of Acetone, 0.4 mole fraction of Methanol and 0.2 Mole fraction of Water can produce
an almost pure Acetone distillate at 1 bar, but the same separation is impossible at 6 bar since this would involve crossing
the distillation boundary. Systems with more azeotropes will, in general, result in more complex residue curve maps. For
example, Figure 4 shows the residue curve map for a mixture of Ethanol, Methyl-Ethyl-Ketone and Water at 1 bar using the Wilson/Ideal
Gas property package. There are three binary azeotropes and one ternary azeotrope yielding a residue curve map with three
distillation regions. A mixture of Acetone, Chloroform and Methanol at 1 bar using the Wilson/Ideal Gas property package (Figure
5) gives, yet again, a more complex residue curve map. Figure 4: Residue Curve Map for Ethanol, MEK and Water at
1 bar. Figure 5: Residue Curve Map for Acetone, Chloroform and Water at 1 bar. Boundary Realtionships Without
a Thermodynamic Model One important aspect to note is that a thermodynamic model is not necessary to make use of residue
curve maps. The knowledge of experimental azeotropes allow us to sketch the structure of the residue curve map and determine,
at least qualitatively, the existence of distillation boundaries. The sketching procedure has been described by Doherty and
Perkins and Doherty and Caldarola (1985). The procedure below is an adaptation of their method and includes an example. This
method will work for many of the common residue curve map that are frequently encountered. However, in some more complex cases,
this method may yield a number of solutions or fail altogether. In these cases where the residue curve map is indeterminate,
you can gain an understanding of the component space by examining the residue curves themselves. Let us assume that you
have the following experimental pure component and azeotropic data: Component 1(mole fraction) Component 2(mole fraction)
Component 3(mole fraction) Temperature( C) 1.00 0.00 0.00 45.5 0.00 1.00 0.00 50.6 0.00 0.00 1.00 52.9 0.52
0.48 0.00 52.1 0.85 0.00 0.15 44.2 0.00 0.61 0.39 41.7 0.38 0.24 0.38 48.8 Table 1: Experimental Pure Component
and Azeotropic Data 1. Begin by drawing and labeling a ternary diagram with all of the experimental data as shown
in Figure 6. Figure 6: Labeled Ternary Diagram. Figure 7: Labeled Ternary Diagram with Completed
Binary Edges. 2. Next, move along each of the binary edges and label the segments between the nodes with an arrow that
points in the direction of increasing temperature. For example, examine the component 1-3 binary edge. Component 1 boils at
a temperature of 45.5 C, the component 1-3 azeotrope boils at 44.2 C (minimum boiling azeotrope), and component 3 boils at
a temperature of 52.9 C. Thus we label the component 1 to component 1-3 azeotrope line segment with an arrow that points toward
component 1 and we label the component 3 to component 1-3 azeotrope line segment with an arrow that points toward component
3. We do the same for the two remaining binary edges, leaving us with a ternary diagram as labeled in Figure 7. 3.
Now we continue by finding the highest and lowest boiling species. The highest boiler is an absolute "stable node":
all residue curves near a stable node point toward and will end at that stable node. The lowest boiler is an absolute "unstable
node": all residue curves near an unstable node begin at and will point away from that unstable node. For example, the
highest boiler is component 3 and the lowest boiler is the component 2-3 azeotrope. Mark the absolute stable node and absolute
unstable node as shown in Figure 8. Figure 8: Labeled Ternary Diagram with Marked Absolute Stable and Unstable
Nodes. Figure 9: Labeled Ternary Diagram with Marked Vertices. 4. Next examine each of the three pure component
vertices and determine their type (unless their type is already known). Look at the arrows on each side of the pure component
vertex. If both arrows point towards the vertex, then the vertex is a stable node. If both arrows point away from the vertex,
then the vertex is an unstable node. If one arrow points towards and one arrow points away from the vertex, then the vertex
is a "saddle". For example, look at the pure component 1 vertex: one arrow points towards and one arrow points away
from the vertex. Thus the pure component 1 vertex is a saddle. Mark the pure component vertices on the ternary diagram with
their types as shown in Figure 9. 5. Now determine the type of the ternary azeotrope (if one exists) by looking at the
possible ternary azeotrope connections. If there are at least two higher boiling and two lower boiling connections possible,
then the ternary is a saddle. Otherwise, it is a node (specific type is determined by comparing its boiling temperature with
the other species). For our example, there are two higher boiling species (pure component 3 and component 1-2 binary) and
two lower boiling species (component 1-3 and component 2-3 binary azeotropes), so the ternary azeotrope is a saddle. 6.
If the ternary azeotrope is a saddle, then connect the ternary azeotrope to two higher boiling species and two lower boiling
species.. Label the connections with arrows that point in the direction of increasing temperature as in Figure 10. The residue
curve map is now complete. Figure 10: Completed Residue Curve Map. If the ternary azeotrope is a stable
node, then connect it to each lower boiling binary azeotrope and each lower boiling pure component. Label each connection
with an arrow that points towards the species that has the higher boiling temperature (increasing temperature). 7. Finally,
if you are left with any binary azeotropes that still have possible connections left, then connect these binary azeotropes
to other unconnected binary azeotropes and pure components. Occasionally we can arrive at several probable residue curve
maps. Some tips for sketching are: 1. Distillation boundaries cannot cross. 2. Only one unstable node and one stable
node can exist within a single distillation region. 3. Pure component saddles cannot make any interior connections. 4.
Binary saddle azeotropes can make a maximum of one interior connection. 5. Ternary saddle azeotropes must make four connections
(two higher boiling, two lower boiling). 6. Unstable nodes cannot connect to any lower boiling species. 7. Stable
nodes cannot connect to any higher boiling species. Thermal Maps and Bubble Point Surfaces Another way of looking
at the fabric of the composition and temperature space is by using thermal maps and bubble point surfaces. A thermal map is
a ternary diagram where color is added to denote the bubble point temperature throughout the component space. A bubble point
surface is a three dimensional version of the thermal map: not only is color used to distinguish between temperature changes,
but actual temperature valleys, ridges and slopes can be observed. Both of these methods aid in the understanding of the component
space on a whole since changes from one area to another are readily visible. The thermal maps and bubble point surfaces that
follow were generated by performing a three phase bubble point flash over a mesh of points separated by a 0.01 mole fraction,
beginning with pure components. Thus, 5151 points were examined in each case. The first thermal map and bubble point surface
shown are in Figures 11 and 12. These figures show a near ideal mixture of n-Pentane, n-Hexane and n-Heptane at 1 bar using
the Wilson/Ideal Gas property package. Figures 11 and 12: Thermal Map (left) and Bubble Point Surface
(right) for n-Pentane, n-Hexane and n-Heptane at 1 bar. Note the regular pattern and near constant slope in Figures 11
and 12 and compare these figures to Figures 13 and 14. Figures 13 and 14 are the thermal map and bubble point surface for
a mixture of Acetone, Chloroform and Methanol at 1 bar using the NRTL/Ideal Gas property package. Figures
13 and 14: Thermal Map (left) and Bubble Point Surface (right) for Acetone, Chloroform and Methanol at 1 bar. Simply by
looking at Figures 13 and 14, we can see that the behavior of this system is far from the ideal behavior in Figures 11 and
12. The Acetone, Chloroform and Methanol mixture contains three binary azeotropes (two minimum boiling, one maximum boiling)
and one ternary azeotrope (saddle). The binary azeotropes are readily identified in the figures and a trained eye quickly
spots the ternary. A system of Ethanol, Methyl-Ethyl-Ketone, and Water at 1 bar and using the NRTL/Ideal Gas property
package gives a particularly interesting thermal map and bubble point surface (Figures 15 and 16). Figures
15 and 16: Thermal Map (left) and Bubble Point Surface (right) for Ethanol, Methyl-Ethyl-Ketone and Water at 1 bar. The
most evident feature in these figures is the minimum boiling ternary azeotrope (unstable node) that is easily identified in
both figures. Note, too, the extremely steep sloping surface near the pure water stable node. Characteristics such as this
can be quickly and easily observed and applied to the operation and control of unit operations. This is possible only because
we have gone past looking at only a few special points in the component space to looking at the space in its entirety.
As a final example of thermal maps and bubble point surfaces, Figures 17 and 18 show a system of Ethanol, Benzene and
Water at 1 bar using the NRTL/Ideal Gas property package. Figures 17 and 18: Thermal Map (left) and
Bubble Point Surface (right) for Ethanol, Benzene and Water at 1 bar Here, we can observe a two liquid phase region as
a relatively flat plateau. As soon as we emerge from this heterogeneous region, the bubble point temperatures steeply increase,
being steepest near the water stable node. Bibliography Doherty, M. F.; Perkins, J. D.; Chem. Eng. Sci., 1978a, 33,
281. Doherty, M. F.; Perkins, J. D.; Chem. Eng. Sci., 1978b, 33, 569. Doherty, M. F.; Perkins, J. D.; Chem. Eng. Sci.,
1979a, 34, 1401. Doherty, M. F.; Perkins, J. D.; Inst. Chem. Eng. Symp. Ser., 1979b, 56, 4.2/21. Doherty, M. F.; Caldarola,
G. A.; Ind. Eng. Chem. Fundam., 1985, 24, 474. Fidkowski, Z.T.; Malone, M. F.; Doherty, M. F.; Comp. Chem. Eng., 1993,
17, 1141. Marsh, K.; Satyro, M. A.; Hyprotechs International Users Conf., 1994. Mathias, P. M.; Copeman, T. W.; Fluid
Phase Equilib., 1983, 13, 91. Satyro, M. A.; Trebble, M. A.; Submitted for publication in Fluid Phase Equilib., 1994.
|